TOWARDS A TYPOLOGICAL CLASSIFICATION OF FALSE FRIENDS

(SPANISH-ENGLISH)
RUBN CHACN BELTRN
UNED
ABSTRACT. Cognate words in two or more languages have a common origin
because of their diachronic relationship and, as a result, they share some sort of formal
and/or semantic affinity. Cognate words can facilitate the foreign language learning
process; they have similar meanings and, therefore, they can support the acquisition
and/or learning of a non-native language. However, these words can also have a
deceptive meaning as a result of semantic change and dissimilar development in two
languages, i.e., they may be deceptive cognate words or false friends. False friends are
especially problematic for language learners as they tend to overgeneralize and assume
they know the meaning of these words, which are actually misleading. In this paper a
taxonomy of cognate words containing six different types is put forward. The
classification is based on semantic and formal criteria (morphological and phonological)
and aims at providing a common framework for the analysis of cognate words which will
help in the development of specific teaching and learning strategies.
KEYWORDS. Vocabulary learning, crosslinguistic influence, false friends, cognate words.
RESUMEN. Las palabras cognadas en dos o ms lenguas son aquellas que tienen
un origen comn dada su relacin desde un punto de vista diacrnico y, como conse-
cuencia de ello, comparten afinidades formales y/o semnticas. Los cognados pueden
favorecer el proceso de aprendizaje de una lengua extranjera, pues tienen significados
similares y as fomentan la adquisicin y/o el aprendizaje de la lengua no nativa. Sin
embargo, estas palabras tambin pueden tener significados ambiguos como resultado
de un cambio semntico y una evolucin diferente en ambas lenguas, esto es, pueden
ser cognados engaosos o falsos amigos. Los falsos amigos resultan especialmente
problemticos para los aprendices de lenguas dado que stos suelen generalizar y asu-
mir que conocen el significado de estas palabras, que en realidad resultan falaces. En
este artculo se presenta una taxonoma de palabras cognadas con seis categoras dife-
rentes. La clasificacin se fundamenta en criterios semnticos y formales (morfolgi-
cos y fonolgicos) y pretende proporcionar un marco comn para el anlisis de
palabras cognadas que contribuya al posterior desarrollo de estrategias especficas de
enseanza y aprendizaje.
PALABRAS CLAVE. Aprendizaje de vocabulario, influencia interlingstica, falsos amigos, cognados.
29
RESLA 19 (2006), 29-39
1. INTRODUCTION
Cross-linguistic influence and its consequences for the teaching and learning of non-
native languages has interested researchers in the last decades, who have concentrated
mainly on the analysis of evidence of L1 traces in L2 phonetic, syntactic or morphological
production. Nevertheless, as Odlin (2003: 437) states, [] anyone seeking to understand
transfer itself in all its manifestations needs to try to become familiar with a wide range of
linguistic research; neglecting to do so can result in making claims that do not square with
the available evidence (as has happened fairly often). More recently, studies have
examined how cross-linguistic evidence affects most linguistic subsystems such as
pragmatics, semantics, syntax, morphology, phonology, phonetics, orthography, etc., not
only in the second but also in the third or subsequent language
1
.
According to Cenoz (2001, 2003) and Kellerman (1983), cross-linguistic influence
tends to be present in language learners production when the mother tongue and the target
language are typologically related. Spanish is a direct descendant of Latin. Even though
English is a Germanic language it has been extensively influenced by Latin at various times,
especially during the Middle English and Early Modern English periods. It can be said that
one of the major sources of loanwords into English, together with French and Scandinavian,
has been Latin. The relationship of English and Spanish with Latin has occasioned a non-
parallel lexical development and thus we can find in both languages many words with a
common Latin origin which nevertheless have evolved differently, giving rise to uneven
correspondences. What is more, since English and Spanish are in contact in many parts of
the world, false friends between these two languages continue to appear.
Cross-linguistic influence in second language acquisition has been related to
language proficiency (Ringbom 1987; Mhle 1989; Poulisse 1990; Cenoz 2001); less
proficient learners generally transfer more elements from their first language than those
who have a high proficiency. This phenomenon can occur with Spanish learners of
English but it should be noticed that the existence of false friends in specialised jargon
is also frequent given the fact that English often incorporates technical words of Latin
origin. Furthermore, the presence of false friends in proficient language users such as
translators, language teachers, journalists, etc. is not to be underestimated because they
are often difficult to identify.
In second language teaching, the historical relationship between the languages in
contact, that is, the L1 (First language) and the L2 (Second language) has generally been
analyzed from two main perspectives as far as the lexical component is concerned: first, as
an element that favors or facilitates the language learning process, depending on the
linguistic closeness (positive transfer) or, second, as a determining factor that hinders the
course of learning, as a result of the interferences caused by the unwanted similarities or
differences between the languages involved (negative transfer). Some studies have been
carried out in order to analyze the facilitating effects of cognate words recognition when
reading in an L2. Authors such as Moss (1992) point out that, in the cases where the L1 and
the target language are historically related and share some helpful similarities, language
RUBN CHACN BELTRN
30
learners should be systematically trained to take advantage of cognate words and thereby
enhance their reading skills and their global understanding of the text.
Ringbom (1986) did some research on the overall proficiency level attained by
Swedish speaking and Finish-speaking EFL (English as a Foreign Language) students in
Finland, where both languages coexist in some parts of the country. These two languages
belong to rather dissimilar language families: Germanic and Finno-Ugric. Ringbom found
a marked tendency on the part of Swedish L1 speakers to outperform Finish L1 speakers
in EFL, both groups having received similar instruction. It was perceived that Swedish-
speaking learners showed a certain tendency to make interlingual identifications probably
because of the similar psycholinguistic routines available in English and Swedish. As far
as reading comprehension is concerned, Ringbom (ibid.) indicates that if the L2 is closely
related to the L1, the language learner will benefit from the existence of cognate words,
given the fact that both, recognition and understanding of these words is less demanding
than completely alien words. In fact, many of these words are not eventually learned but
the formal similarity, especially in writing, helps the language learner to understand the
text and to accomplish a smooth reading but, conversely, there is little psycholinguistic
processing. Rather, unconsciously, the language learner tends to consider cognate words as
a help for his reading which do not require special attention. So, Ringbom (1992)
introduces the idea of potential knowledge to refer to the learners knowledge or familiarity
with a word or grammar construction which, in fact, has not been seen before in the L2. It
goes without saying that the closer the typological proximity between languages, the more
chances the language learner has to find instances of this potential knowledge, at least as
far as receptive skills are concerned, i.e., listening and, especially, reading. Whereas the
absence of cognate words between the L1 and the L2 considerably reduces the amount of
familiar vocabulary that the language learner has access to, and the range between
active and passive vocabulary diminishes considerably (Ringbom, 1992). Ringboms
research centers on two languages which are rather close from a morpho-syntactic point of
view, i.e. English and French.
Ringbom (2001) divides various types of lexical transfer into two main blocks,
transfer of form and transfer of meaning. The former type comprises mainly complete
language switches and the use of deceptive cognates, which may be partially or totally
deceptive. The latter refers to other instances of lexical transfer like calques and
semantic extension on the basis of patterns in other languages (Ringbom, 2001). So, as
we can see in Ringboms classification, deceptive cognate words are broadly classified
into partial and total without further detail. Furthermore, this classification defines false
friends, or deceptive cognates, as form-based transfer instead of as meaning-based
transfer as suggested in Postigo Pinazo (1997).
2. FALSE FRIENDS IN EFL
In the same way as cognate words, or rather true cognates, bring about undeniable
help in the development of some linguistic abilities in learners with specific linguistic
TOWARDS A TYPOLOGICAL CLASSIFICATION OF FALSE FRIENDS (SPANISH-ENGLISH)
31
backgrounds, there are also cognate words with a deceptive meaning, often known as
deceptive cognates or false friends, and which may entail a learning difficulty. False
friends are not exceedingly common in Spanish-English but in some contexts
2
they
represent a true learning problem as they become rather frequent. So, although their
occurrence is not too high, it is not so scarce either as not to require special attention on
the part of researchers and language teachers. Arnold (1992) confirmed that not knowing
the meaning of some false friends in reading activities is potentially more dangerous that
not knowing the meaning of unfamiliar words, because in the former case students
usually try to infer the meanings of those familiar words without checking them. It
should also be taken into consideration that when a language learner misunderstands a
false friend, it is very improbable that s/he will realise the mistake unless negative
evidence is provided by means of explicit information (Lightbown and Spada 1993).
In relation to the degree of difficulty in learning false friends, this may be
determined by two inherent features, that is, if they are total false friends or if they are
only partially deceptive. The first group comprises those false friends which have an
utterly different meaning in both languages (e.g.: Engl. terrific; Span. estupendo),
whereas the second group, partial false friends, refers to polysemous words, one of
whose meanings is a false friend while another (or others) is a true cognate word (e.g.:
Engl. approve means dar el visto bueno or dar su aprobacin, whereas Span. aprobar
also means pasar una prueba o examen). It is precisely this last type of false friends that
can be most confusing for EFL students.
Following this pedagogic perspective, Frantzen (1998) distinguishes two factors that
may determine the degree of false friends difficulty: intrinsic and extrinsic factors. Intrinsic
factors emerge from the confusing character of false friends per se, for example the fact that
some of them have a deceptive meaning in all circumstances whereas some others may not
be deceptive in certain contexts. This fact leads EFL students to a situation of considerable
uncertainty because, in some cases they are not confident if their meaning is misleading in
all instances (reliably false) (e.g.: Span. carpeta Engl. carpet, Span. fbrica Engl.
fabric) or if, on the contrary, in some cases they can be, in fact, true cognates (unreliable
false) (Span. introducir English to introduce s.o. to s.o. else, but it also means bring up
a topic). Another intrinsic factor lies in the fact that the semantic field in which they tend
to appear may also overlap; so, false friends may have meanings in both languages which
are within the same semantic field (e.g.: English constipated and Span. constipado, these
words are both within the semantic field of health). The third intrinsic factor that Frantzen
points out is that students may show a tendency to overgeneralize as a result of the large
number of cognate words (true friends) that they may come across, perhaps induced by the
teacher or the teaching methodology.
Extrinsic factors arise from the different types of contradictory input that students
may be exposed to. On some occasions, and as can be seen in other research articles cited
above, language learners are sometimes encouraged to make use and take advantage of
true cognates, without being warned of the existence of false friends. This practice may
give rise to a situation of frustration on the part of the language learners as they may
RUBN CHACN BELTRN
32
perceive that it is a problem-inducing and confusing strategy. As a result of this
circumstance, it is likely that students develop a feeling of dissatisfaction as they
perceive that this inferring strategy sometimes works and sometimes does not.
Students may avoid cognate words (true friends) because they suspect that they do not
really mean what they look. Frantzen indicates another extrinsic factor which is less
representative in EFL contexts but rather plausible in an L2 learning context. Some false
friends have become commonplace in certain Spanish-speaking communities within the
US, (e.g.: marqueta market mercado). She also refers to the number of potential
mistakes within the EFL classroom context, the lack of systematicity in media
translations, and the oversimplification of lexicographers when writing dictionaries.
3. TOWARDS A TYPOLOGICAL CLASSIFICATION OF FALSE FRIENDS
As regards the classification of true cognate words and false friends, not many far-
reaching advances have been attempted so far. Nevertheless, some individual research
projects have tried to organize them, with different purposes in mind, but there is not a
standardized typology comprising formal and semantic criteria at the same level of
analysis.
Moss (1992) classified a written corpus of false friends according to the following
criteria: (a) word length (number of letters); (b) number of differences between English
and Spanish words; (c) the proportional difference (i.e., the proportion of differences
with respect to word length); (d) part of speech; (e) whether the initial and final letters
are the same or different; (f) the number of vowel differences (these are divided into
addition, deletion and change); (g) the number of consonant differences (divided into
addition, deletion, and change); and, (h) specific letter differences.
Postigo Pinazo (1997) collected a corpus of over 1,800 potentially misleading
words (Spanish-English) and classified them into the following four categories: (a)
phonetic false friends; (b) graphic false friends; (c) false friends derived from loanwords;
and, (d) semantic false friends, which could be subdivided into total or partial. This
classification seems rather imprecise because it does not attempt to find connections
between three basic levels of analysis, the semantic, morpho-syntactic and phonological
ones, which certainly overlap. In addition to this, no implications for language learning,
or for any other applied aspect of language use like, for instance, translation, can be
drawn from this classification. Postigo Pinazo (1997) gathered a number of Spanish-
English false friends and made reference to their linguistic form, their origin and their
meaning but dealt with these three categories as mutually exclusive, when they are, in
fact, interrelated since all false friends regardless of their form or origin are semantic.
The present study adopts a pedagogic perspective as false friends are considered a
psycholinguistic learning problem that requires special consideration within the
language learning classroom. So, false friends are not analyzed with regards to a
historical perspective, their diachronic background, or their evolution through time, but
simply as potential learning problems regardless of their origin.
TOWARDS A TYPOLOGICAL CLASSIFICATION OF FALSE FRIENDS (SPANISH-ENGLISH)
33
Given the lack of conclusive research and broad categorizations of cognate words,
with either reliable or unreliable similarities, the classification below has been
developed. The CCVF (Clasificacin de Cognados Verdaderos y Falsos) intends to help
in the taxonomy of certain groups of cognate words in the hope that this typology can
facilitate and pave the way for future research in the fields of Foreign Language
Teaching and Learning. However, it should be explained that the following graphic
describes a complicated linguistic feature and whenever such an attempt given the
complexity of the linguistic system is made, some other linguistic aspects are inevitably
simplified. For instance, the fact that false friends may vary depending on the language
variety (British and American English) is not taken into account in the present
classification. It should also be borne in mind that cognate words are language specific
for each pair of languages, and on this occasion it is only valid for English and Spanish.
A representation of the cognate words mental lexicon in the L1 and the L2.
RUBN CHACN BELTRN
34
Graphic 1. A typology of Spanish-English true cognates and false friends (CCVF)
1 = True Cognates: Phonetic.
2 = True Cognates: Graphic.
3 = Partial False Friends: Phonetic.
4 = Total False Friends: Phonetic.
5 = Partial False Friends: Graphic.
6 = Total False Friends: Graphic.
The graphic above represents a schematization that may help to classify cognate
words according to three variables. Firstly, if they are true cognates or false cognates
(false friends). Secondly, if they are graphic or phonetic cognate words, that is, if their
pronunciation evokes the pronunciation of a word in the L1 (phonetic) (e.g.: Engl.
television /telIvIZn/ and Spa. televisin /teleBIsjon/), or if their written form recalls
a word in the L1 even though the pronunciation may differ utterly in both languages
(graphic) (Engl. arena /ri>n/; Span. arena /aRena/). Thirdly, in the case of false
friends, if they are partial or total from a semantic point of view, i.e, total false friends
entail the univocal relationship between a meaning A in the L1, and a meaning B in the
L2, whereas partial false friends involve a meaning A in the L1 but some other
meanings in the L2 (A, B, C, D, etc.). See table 1 for a detailed analysis of cognate word
difficulty.
Meaning Spelling Pronunciation Degree of difficulty
Type 1 + + + Nil
Type 2 + + Low
Type 3 + / + + High
Type 4 + + Medium
Type 5 + / + Very high
Type 6 + Medium
Table 1
According to the CCVF there are six possible types of cognate words; two of them
qualify as true cognates and the remaining four as false friends.
A. TRUE COGNATES: PHONETIC
E.g.: Engl. laboratory /lbrtri/ = Span. laboratorio /laBoRatoRjo/
vocabulary /vkQbjUlri/ = Span. vocabulario /bokaBulaRjo/.
These cognate words may make a valuable contribution to the process of learning
a non-native language because they can aid EFL learners given the semantic,
orthographic and phonological similarities with the L1. Whenever a language learner
TOWARDS A TYPOLOGICAL CLASSIFICATION OF FALSE FRIENDS (SPANISH-ENGLISH)
35
>
hears a word included in this category, s/he can easily identify the word given the
pronunciation similarities with the L1 (See table 2 for further examples). It would need
to be specified that these words showing a similarity in their oral form, additionally
exhibit useful analogy in the written form.
Examples
Type 1 Effective, consonant, perfect, interesting.
Type 2 Isle, horizon, ocean, triangle.
Type 3 Fabric, suspend, support, faculty (AmE), disorder, involve.
Type 4 Abrupt, relevant, evidence, sympathy, sensible, pamphlet, edit.
Type 5 Coincide, decent, education, spade, enroll, err.
Type 6 Audacious, signature, vegetables, journal, scholar, lantern, vacation.
Table 2
B. TRUE COGNATES: GRAPHIC
E.g.: Engl. horizon /hraIzn/ = Span. horizonte /oRiTon5te/
Engl. ocean /USn/ = Span. ocano /oTeano/.
Words included in this category are very helpful for the language learner from a
semantic point of view, too. In this case, the similarity embraces the written form of the
word but not so much the pronunciation and, therefore, the language learner may fail to
identify the word in listening and will probably be unsuccessful in pronouncing it
appropriately if a model is not provided (See table 2 for further examples).
C. PARTIAL FALSE FRIENDS: PHONETIC
E.g.: Engl. (to) attend /tend/ = Span. asistir a (to be present)
Span. prestar atencin (to listen carefully)
Span. atender /aten5deR/ (to pay attention)
(to) attend to = Span. ocuparse de
Span. atender (a shop-assistant)
At this level of analysis, the semantic component is introduced as a component in
the classification of cognate words, and the partial/total variable refers to that
characteristic. Partial false friends have a main meaning in one of the languages but may
represent a number of different meanings in the other language. Total false friends,
however, show a clear semantic difference in both languages while marinating a
correspondence between one word and one meaning. At the phonetic level, the affinity
between the languages is strong enough to let the language learner perceive the aural
similarity. (See table 2 for further examples).
RUBN CHACN BELTRN
36
D. TOTAL FALSE FRIENDS: PHONETIC
E.g.: Engl. (to) assist /sIst/ = Span. ayudar
Span. asistir /asistiR/ (help)
Span. asistir (to attend)
Total false friends cause a learning problem given the lack of semantic
correspondence. In any case, this absence of equivalence is one-to-one in the L1 and the
L2, and that determines that these words do not become excessively difficult for the
language learner. In this case the cognitive processing in learning the word requires that
the language learner assign a new meaning to a word that seems to have another
meaning. Pronunciation does not present any difficulty to the language learner. (See
table 2 for further examples)
E. PARTIAL FALSE FRIENDS: GRAPHIC
E.g.: Engl. career /krI/ = Span. carrera profesional (E.g.: carrera judicial)
Span. trayectoria profesional
Span. carrera /kareRa/ (universitaria)
Engl. agenda /dZend/ = Span. orden del da
Span. agenda /axenda/ (programa)
Span. libreta (diary)
Just as in type 3, words included within this category present a hindrance related
to the lack of semantic equivalence between the L1 and the L2. What is more, while
there is no major difficulty in the written recognition of the word, a serious problem
arises in the aural identification and oral production of the word. (See table 2 for
further examples).
F. TOTAL FALSE FRIENDS: GRAPHIC
E.g.: Engl. qualifications /kwlIfIkeISnz/ = Span. requisito, antecedentes laborales
Span. calificacin /kalifikaTjon/
(grades)
Engl. lecture /lektS/ = Span. charla, conferencia, clase universitaria
Span. lectura /lektuRa/
Parallel to type 4, type 6 false friends present an unequivocal dissimilarity in
meaning between the L1 and the L2. In relation to their formal written identification they
are not exceedingly complex but their pronunciation is often not so obvious to the
language learner. (See table 2 for further examples).
TOWARDS A TYPOLOGICAL CLASSIFICATION OF FALSE FRIENDS (SPANISH-ENGLISH)
37
4. CONCLUSIONS
The taxonomy described above may have a practical application in the learning of
foreign languages, particularly in the learning of English as a Foreign Language. Once
this framework for the classification of cognate words has been developed, the next step
which is beyond the aims of this paper would be to classify most frequent cognate
words (Spanish-English) according to this common framework (see table 2 for an
instance). Arising out of this taxonomy different teaching/learning tasks and exercises
can be developed according to the word type within the CCVF and their perceived
degree of difficulty, which would certainly suggest that it would be wise to introduce
different types of false friends at different proficiency levels. Later, different teaching
techniques and methodologies could be elaborated for the teaching of false friends
following various types of focus on form which have proven to be remarkably effective
in teaching false friends (Chacn Beltrn, 2001) in a classroom context. A further
implication of this study is that the CCVF could be applied to other pairs of languages,
taking into account that it will prove more efficient in historically related languages
which have evolved differently.
Odlin (2003: 478) points out that [] there does not yet exist any comprehensive
theory of language transfer - and the appearance of one any time soon seems unlikely.
In the meantime domain-specific studies such as the one presented here seem to be the
only way to research crosslinguistic influence.
NOTES
1. See Cenoz, Hufeisen, and Jessner (2001).
2. In academic English, for instance, many false friends are to be found.
REFERENCES
Arnold, J. 1992. Transferencia positiva y negativa en la lectura del ingls. Lingstica
aplicada al aprendizaje del ingls. Eds. A. Bruton and J. Arnold. Sevilla:
Universidad de Sevilla.
Cenoz, J. 2001. The effect of linguistic distance, L2 status and age on cross-linguistic
influence in third language acquisition. Cross-linguistic Influence in Third Language
Acquisition. Eds. J. Cenoz, B. Hufeisen, and U. Jessner. Clevedon: Multilingual
Matters. 8-20.
Cenoz, J. 2003. The role of typology in the organization of the multilingual lexicon.
The Multilingual Lexicon. Eds. J. Cenoz, B. Hufeisen, and U. Jessner. Kluwer
Academia Publishers. 103-116.
Cenoz, J., Hufeisen, B., and U. Jessner, eds. 2001. Cross-linguistic Influence in Third
Language Acquisition. Clevedon: Multilingual Matters.
RUBN CHACN BELTRN
38
Chacn Beltrn, R. 2001. La enseaza de vocabulario en ingls como L2: el efecto del
nfasis en la forma lingstica en el aprendizaje de cognados falsos. Tesis Doctoral.
Universidad de Sevilla. [Internet document available at http://fondosdigitales.us.es/
thesis/thesis_view?oid=458]
Frantzen, D. 1998. Intrinsic and extrinsic factors that contribute to the difficulty of
learning false cognates. Foreign Language Annals 31(2): 243-254.
Kellerman, E. 1983. Now you see it, now you dont. Language Transfer in Language
Learning. Eds. S. Gass and L. Selinker. Rowley, MA: Newbury House.
Lightbown, P. and N. Spada. 1993. How Languages are Learned. Oxford: Oxford
University Press.
Mhle, D. 1989. Multilingual interaction in foreign language production. Interlingual
Processes. Eds. H.W. Dechert and M. Raupach. Tbingen: Gunter Narr. 179-194.
Moss, G. 1992. Cognate recognition: Its importance in the teaching of ESP reading
courses to Spanish speakers. English for Specific Purposes 11: 141-158.
Odlin, T. 2003. Cross-linguistic influence. The Handbook of Second Language
Acquisition. Eds. C. J. Doughty and M. H. Long. Malden, MA: Blackwell Publishing.
Poulisse, N. 1990. The Use of Compensatory Strategies by Dutch Learners of English.
Dordrecht: Foris.
Postigo Pinazo, E. 1997. Estudio contrastivo de los falsos amigos en ingls y en espaol.
Mlaga: Servicio de Publicaciones Universidad de Mlaga. [Microficha]
Ringbom, H. 1986. Crosslinguistic influence and the foreign language learning process.
Crosslinguistic Influence in Second Language Acquisition. Eds. E. Kellerman and M.
Sharwood Smith. New York: Pergamon Press.
Ringbom, H. 1987. The Role of the First Language in Foreign Language Learning.
Clevedon: Multilingual Matters.
Ringbom, H. 1992. On L1 transfer in L2 comprehension and L2 production. Language
Learning 42: 85-112.
Ringbom, H. 2001. Lexical transfer in L3 production. Cross-linguistic Influence in
Third Language Acquisition. Eds. J. Cenoz, B. Hufeisen, and U. Jessner. Clevedon:
Multilingual Matters. 59-68.
TOWARDS A TYPOLOGICAL CLASSIFICATION OF FALSE FRIENDS (SPANISH-ENGLISH)
39